Emphasise table captions

Author: matthew-brett

advanced/advanced_numpy/index.md

@@ -158,7 +158,7 @@ block.
 
 {class}`dtype` describes a single item in the array:
 
-::: {list-table} Dtypes
+::: {list-table} **Dtypes**
 
 - - type
   - **scalar type** of the data, one of:

advanced/mathematical_optimization/index.md

@@ -300,7 +300,7 @@ gradient, that is the direction of the _steepest descent_.
 
 +++
 
-::: {list-table} Fixed step gradient descent
+::: {list-table} **Fixed step gradient descent**
 
 - - **A well-conditioned quadratic function.**
 
@@ -345,7 +345,7 @@ Also, it clearly can be advantageous to take bigger steps. This
 is done in gradient descent code using a
 [line search](https://en.wikipedia.org/wiki/Line_search).
 
-::: {list-table} Adaptive step gradient descent
+::: {list-table} **Adaptive step gradient descent**
 
 - - A well-conditioned quadratic function.
 
@@ -409,7 +409,7 @@ it cross the valley. The conjugate gradient solves this problem by adding
 a _friction_ term: each step depends on the two last values of the
 gradient and sharp turns are reduced.
 
-::: {list-table} Conjugate gradient descent
+::: {list-table} **Conjugate gradient descent**
 
 - - An ill-conditioned non-quadratic function.
 
@@ -773,7 +773,7 @@ See [compare optimizers](compare-optimizers-eg).
 
 :::
 
-::: {list-table} Rules of thumb for choosing a method
+::: {list-table} **Rules of thumb for choosing a method**
 
 - - Without knowledge of the gradient
 

intro/intro.md

@@ -48,7 +48,7 @@ Valentin Haenel_
 
 ### How does Python compare to other solutions?
 
-::: {list-table} Compiled languages (C, C++, Fortran ...)
+::: {list-table} **Compiled languages (C, C++, Fortran ...)**
 
 - - Pros
   - Very fast. For heavy computations, it’s difficult to outperform these
@@ -60,7 +60,7 @@ Valentin Haenel_
 
 :::
 
-::: {list-table} Matlab scripting language
+::: {list-table} **Matlab scripting language**
 
 - - Pros
   - - Very rich collection of libraries with numerous algorithms, for many
@@ -76,7 +76,7 @@ Valentin Haenel_
 
 :::
 
-::: {list-table} Julia
+::: {list-table} **Julia**
 
 - - Pros
   - - Fast code, yet interactive and simple to read and write.
@@ -87,7 +87,7 @@ Valentin Haenel_
 
 :::
 
-::: {list-table} Other scripting languages: Scilab, Octave, R, IDL, etc.
+::: {list-table} **Other scripting languages:** Scilab, Octave, R, IDL, etc.
 
 - - Pros
   - - Open-source, free, or at least cheaper than Matlab.
@@ -101,7 +101,7 @@ Valentin Haenel_
 
 :::
 
-::: {list-table} Python
+::: {list-table} **Python**
 
 - - Pros
   - - Very rich scientific computing libraries

packages/scikit-image/index.md

@@ -46,7 +46,7 @@ Images are NumPy's arrays `np.ndarray`
 
 +++
 
-::: {list-table} Terms
+::: {list-table} **Terms**
 
 - - Pixels
   - array values: `a[2, 3]`
@@ -77,7 +77,7 @@ Python installations, as well as in most Linux distributions. Other
 Python packages for image processing & visualization that operate on
 NumPy arrays include:
 
-::: {list-table} Other packages for working with images
+::: {list-table} **Other packages for working with images**
 
 - - {mod}`scipy.ndimage`
   - For N-dimensional arrays. Basic filtering, mathematical morphology,
@@ -91,7 +91,7 @@ NumPy arrays include:
 
 Some powerful C++ image processing libraries also have Python bindings:
 
-::: {list-table} C++ libraries with Python bindings
+::: {list-table} **C++ libraries with Python bindings**
 
 - - [OpenCV](https://docs.opencv.org/4.x/d6/d00/tutorial_py_root.html)
   - A highly optimized computer vision library with a focus on real-time
@@ -117,7 +117,7 @@ The library contains predominantly image processing algorithms, but
 also utility functions to ease data handling and processing.
 It contains the following submodules:
 
-::: {list-table} Scikit-image submodules
+::: {list-table} **Scikit-image submodules**
 
 - - {mod}`skimage.color`
   - Color space conversion.

packages/scikit-learn/index.md

@@ -500,7 +500,7 @@ Scikit-learn strives to have a uniform interface across all methods, and
 we’ll see examples of these below. Given a scikit-learn _estimator_
 object named `model`, the following methods are available:
 
-::: {list-table} Estimator interfaces
+::: {list-table} **Estimator interfaces**
 
 - - All Estimators
   - - `model.fit()` : fit training data. For supervised learning